I think this actually speaks quite nicely to Bergson's point here re: what I referred to as a desubstantialization of time: to the degree that every space-time block is relative to the trajectory of it's localising elements, what is missing or simply untheorized is precisely the passage from one trajectory to another. In other words what is missing, or rather, what is simply assumed is time itself. Time is 'given': given this space-time trajectory, that is the corresponding space-time block. But the passage of time itself is precisely what takes place 'before' (logically speaking) STR 'kicks in', as it were.
I find a lot to agree with, but I wanna place some extra emphasis on the strengths of the theory. I believe it is informative to a naturalistic metaphysics - at least insofar as it rules out various ontological postulates.
For probably the first time since the old PF, I'll be using my own idiosyncratic concepts a lot. So bear with me.
The only thing I find disagreeable in this is that passage from one trajectory to another is internalised through the equivalence principle. Say we consider two frames of reference whose equations of motion are given by sets F1 and F2, there is a relativised form wherein velocity is mapped to relative velocity, accelerations are mapped to relative accelerations and so on - these allow the translation of specific space time trajectories to one another.
I don't think this is a merely epistemic property of SR, I think it's essentially a statement of
relational closure of motion in the regional ontology of SR. Will digress on relational closure to make some sense of it, as it's a term I use when thinking about things like this. Discussing what a regional ontology is will follow.
If someone believes in a transcendent God, they stand in a relation to that God. Specifically X believes that {some set of properties obtain about that God}. If it is really that transcendent God which is related to, how is possible for something 'exterior to being' in substantive senses or 'logically prior to it' in others to relate to any particular? The impossibility of this I term relational closure, relata are always existent in some commensurable way, and moreover there must be an overlap of the modes of being/becoming of the relata in the relation.
This is to say - different trajectories of motion in SR, whatever their sense of time - have common modes of expression, and specifically the relations in this case are the mapping of velocities to relative velocities and coordinate systems to others. There's no external 'space of trajectories' to go to for any regional ontology of SR. I don't mean 'SR gives a good account of aging' or 'SR gives an account which involves human reaction time' or anything to do with evolution, which have their own regionalised notions of time, just that trying to find a sort of movement which isn't described well by SR in its own terms is pretty hard, if not impossible. You can find examples by going close to very dense objects, however, and in these cases GR takes over.
For regional ontology, I refer to the specific mode of being of a designated class of phenomena. The only classes which can have regional ontologies devoted to them are ones with relational closure. I think this is similar to 'creating concepts on a plane of immanence' in Deleuze (or as it's thought of by Daniel Coffeen). Further, developing a regional ontology of a class of phenomena corresponds to the study of how it relates its subunits, and how those relations relate.
It's in this sense that coordinate transformations in SR are not just principles of relations of the objects in SR, they're also the means by which the relations are related. The generation of all of these relations for the regional ontology of SR comes from two principles, the cosmic speed limit of light and the equivalence between relative motion and (Lorentz) coordinate transformations (the equivalence principle). Everything follows from those building blocks.
That said, the passage of time can be said to be presumed in SR and GR insofar as it is primitive in the relations of SR and GR - it's a dimension of a vector space, which when combined with others (space) it produces/models the relativisation of time and space through the category of motion. Not simply in the sense that 'motion is space change over time' or that 'motion is change', but it effects the hows of both - motion
as such, just as time
as such is not part of the regional ontology.
One cannot draw a temporal ontology from it without losing the very thing it ought to account for: time.
In some senses, yes, in some senses no. Heidegger treats regional and fundamental ontology in a similar manner, attempting to ground regionalised things like moods and language in more primordial things like self-direction and the interpretive-as-structure. But I think this is an idealist way of looking at it - as if the ontology of Dasein was derived from but not conditioned by its ontic realisations. Similarly, SR does place a few constraints on what an ontology of space, time and motion should look like; how it deals with motion
as such, time
as such and space
as such. By
as such here I still mean in a somewhat constrained sense relating to physical properties and material flows.
(1) Time
itself should be considered as something which can interact with its own unfolding, considering that motion may change its
rate of unfolding.
(2) Space
itself should be considered as something inseparable from time and vice versa, otherwise space-time curvature doesn't make sense.
(3) Motion
itself should be considered as more than an analytic composite of time and space, since it effects the unfolding of its constitutive objects
and itself. It is both a relational category of time and space as well as a phenomenon effecting both.
(1),(2),(3) together mark space,time and motion as of equivalent logical priority. Ontologies should treat them as distinct but inseparable.
There are also more specific conclusions that can be drawn:
(A) Time
as such cannot be subordinated to a notion of event succession, since temporal orders of events are relativised to motion. The event-clock of the universe can be fickle.
(B) Space
as such cannot be subordinated to extension, as time is implicated within it (and vice versa) and these things all change depending on what is analysed. There is a little wrinkle here in terms of treating time as an extensional dimension as well as an index of events, but axiomatically positing the two as the same is what happens when you make it a dimension of the space-time vector space.
Another thing relativity teaches us is that a different regional ontology is needed for the every-day. That is, non-relativistic thinking. We can't ignore the various limiting theorems that reduce GR to SR and SR to Newton. In this realm temporal order is preserved. Our technology can push us past scenarios of the every-day, however, but the effects of it are usually confined (like in the LHC) or very small (like clock differences on transpacific flights). In other words, we can't say that the ontology of GR and SR is a complete account of their contained concepts, even though there is a large region of overlap between their respective 'as such' categories and the theories. They do quite a lot to explain various phenomena in the universe, and so should be treated with respect within their zones of relevance.
Another big thing to note is that experiential time, and the relation of experiential time to physical time, are largely untouched by both.